602: How I’m Using AI To Grow My Store And Courses In Unexpected Ways
Ecom Podcast

602: How I’m Using AI To Grow My Store And Courses In Unexpected Ways

Summary

"By using AI to automate 'bought together' recommendations, one store saw an 18% sales lift in just one day, with expectations to stabilize at 10-20%, showcasing how AI can simplify complex coding tasks and boost revenue in e-commerce."

Full Content

602: How I’m Using AI To Grow My Store And Courses In Unexpected Ways Speaker 1: Welcome back to the podcast, the show where I cover all the latest strategies and current events related to ecommerce and online business. Now, in this episode, I'm sharing the AI changes that I've made to my store and courses that are directly boosting revenue. You'll hear about what's working right now, what surprised me, and how you can apply these ideas to your own business. But before we begin, I want to let you know that tickets for Sellers Summit 2026 are now on sale over at SellersSummit.com. And if you sell physical products online, this is the event that you should be at. Unlike most ecommerce conferences that are filled with high-level fluff and inspirational stories, Sellers Summit is all about tactical, step-by-step strategies you can actually use in your business right away. Every speaker I invite is in the trenches, people who are running their own ecommerce stores, managing inventory, dealing with suppliers, and scaling real businesses. No corporate execs, no consultants. Also, I hate big events, so I intentionally keep it small and intimate. We cap attendance at around 200 people so you can actually have real conversations and connect with everyone in the room. We've sold out every single year for the past nine years and I expect this year to be no different. It's happening April 21st to April 23rd in Fort Lauderdale, Florida. And if you're doing over $250,000 or $1 million in revenue, we also offer a private mastermind for higher level sellers. Right now, tickets are the cheapest they're ever going to be. So if you want in, go over to sellerssummit.com and grab your ticket. Now on to the show. Welcome to The My Wife Quit Her Job Podcast. Today we're going to be covering everything that I've been doing while Toni has been traveling Europe on vacation. Speaker 2: I wasn't on vacation. Unknown Speaker: Let's be clear. Speaker 1: It looked like a vacation to me. I saw your Instagram feed. Speaker 2: Actually, oh my gosh, I had a great podcast topic and I couldn't remember what it was this morning because I'm 52 and I can't remember anything anymore. But we have to do a podcast in the future about can you really be a digital nomad, like still work and travel. Speaker 1: Oh yeah, that's a good one. Speaker 2: Yeah, so that's coming soon, folks. But anyway, yes, I'm curious because I feel like I left and all of a sudden you got super productive. So I was like, am I holding Steve back most days? Because I leave the country and all of a sudden he's built robots and all sorts of things and really cool stuff actually. Speaker 1: You were gone a month, right? Speaker 2: I was gone a month. Speaker 1: You were gone a month. Okay. Let's go over the stuff that's fresh in my head like literally I launched something yesterday and already within one day. According to the stats, it's lifted my sales 18% for just that one day, but I'm optimistic. I'm pretty sure it's not going to stay at 18%, but I'm anticipating 10% to 20%. This has been on my list for a year to do with my store. You know how Amazon has that bought with? Okay, and I've always wanted to implement it, but it is a pain in the butt to implement because you gotta manually put all the bot with stuff together. Speaker 2: Okay. Speaker 1: Or you use code. Okay, so let's start from the beginning because this is a much more tedious problem than I thought originally. So the first thing that I wanted to do was, you know, items that are commonly bought together, statistically figure out what those are. Speaker 2: Right. Speaker 1: And then just display that. That is relatively easy, but I didn't know how to do it. But now that there's AI, I can just have AI write the code. This is the only time I can talk about this geeky stuff. Turns out there's this package called FP Growth. Where you just feed it all of your sales data. Speaker 2: Okay. Speaker 1: And it'll tell you what goes with what, at what confidence level, and at what lift percentage. Speaker 2: Okay. Speaker 1: Right. Speaker 2: So for example, let's say that you're talking that does this. Speaker 1: It is a Python library. Let's say item B is often bought with item A. The confidence is the probability that that's going to happen and then the lift is how much more likely that item is bought with that. You literally have to go through your database and dump out every single order that you've ever gotten and what it's correlated at. And then it creates this chart and then I update that into the database. So now every time you look at a product, it shows what that product is commonly bought with. Nothing fancy, right? Amazon's had this for a long time. The problem with my store is we have like, I don't know, almost a thousand SKUs. And so that isn't actually populated for a lot of the products. Because in order for this to work, there needs to be a lot of products bought with each other. And you know in any typical store, it's like 80-20, right? Only 20% of your products. And so I had tons of similar products for my best sellers, which is great. But then I had this whole library of products that didn't have anything. And I told you the stats earlier, 18% lift. So I want that on every single product. The other problem is you don't even know whether products are being bought with each other because you're displaying them with each other or because people are finding them. It's kind of like a chicken and egg thing. And so what I did, and this wasn't possible before, is I had AI generate me all the similar products for every single product in the library. I didn't know this was possible literally until I tried it last week, but you can feed it an image and then turn the image into math and then compare that image with every single product on your site and find the most similar ones, put them in a database and always have that always bought thing populated. Speaker 2: Interesting. Speaker 1: So now if someone buys like a Battenberg lace handkerchief, the AI finds everything that looks like that Battenberg lace handkerchief and then displays it in the bought with. So now everything is populated. Speaker 2: Yeah, so if there's a product that normally doesn't have a bought with, it's just going to show similar type products below. Speaker 1: Correct. Speaker 2: Yeah, that's smart. I like that. Speaker 1: Also, on my site, I don't know if you've been on Bumble Bee Linens for a while, whenever you click Add to Cart, there's a pop-up that comes up that suggests products for you to add to your cart. Prior to that, I was just using also bought data, but I didn't have a good algorithm. I was just showing something random. That someone has bought before, like among this huge library. I switched it over to a combination of frequently bought plus the similar item. Like if there's not enough similar items, I populate it, I add on to it with the similar products. And that has grown that side probably like double the lift just from doing this. Because my pop-up wasn't working that well before. This is really interesting because Amazon has all this data, but I never had all this data for my site. But you know how underneath the product, there's a box that says you might also be interested in? Amazon has this, and I've had this for a long time. But once again, the products down there were just kind of like a random set of products, and it wasn't populated for everything. And so that lifted tremendously. Speaker 2: Yeah, I can see why because when I think about the Amazon shopping experience, especially with certain types of items, right? Like when I put a book in my cart, I'm very interested to see what other people who like that book also read, right? And they that's a big one, right? Because then it lists And the books always make sense, right? If you're getting a book on marketing, typically people who buy that book buy other marketing books, right? Same thing with clothing. If people buy this shirt, they tend to buy a belt or a pair of pants or things like that. It's kind of putting an outfit together for you inadvertently because of what other people are buying. I think One Click Upsell does this to some, you can do this to some degree. They have some AI tools that help you do that, but it's not, It doesn't always show up right away. Sometimes it's like after the purchase or like further down the purchase cycle as opposed to like when you just add it to your cart or when you're just looking at the product page, right? Because I know on Amazon, it's on the product page, right? Speaker 1: Yeah, it's on the product page. Speaker 2: So I think OneClickUpsell has a little bit of this capability but not to the level that you're talking. Speaker 1: Well, it's funny. I was going to give a lesson in my class about this and whenever I do this now, I look for a Shopify plugin that does it. And you're right. One click upsell kind of does this, but the problem with this is it's computationally intensive. Like that FP growth thing I was talking about, it literally takes probably like a couple of minutes to calculate on my computer when I feed it everything. And then that image similarity, index that turns into math and compares the similarity, that also takes like a couple of minutes for that to happen. Speaker 2: Yeah. Speaker 1: And you can't do that in real time. The only way to do it is to, the way I do it is whenever I add a new product, it automatically regenerates everything. And then once a week, I do the other thing that calculates all the sales and what's bought with each other. Because it takes time, you can't do it on the fly as someone's shopping. Unless like someone has a super fast machine, like maybe I don't have a fast machine, I don't know. You can't do it in real time, otherwise it would bog the site down. And perhaps that's why the tools probably don't do it to the level of accuracy. There's ways to like shortcut the problem. Speaker 2: Yeah. I'll be interested to see like in 30 days where you're at with those numbers. Speaker 1: Yeah, but I was just shocked and like one day it took me longer to code up all the tracking than it did to code up the actual feature itself. Speaker 2: Interesting. Speaker 1: But yeah, just in one day. So we'll see. Yeah, we'll revisit this and maybe in 30 days. Speaker 2: I do think to caution people, I think this is great and I think this is a great thing to have, but I think anytime you make a big change on a website, you're going to see a very high lift initially, right? And then it's going to level out. I think you'll still see a lift for sure. But what I've seen people do is get that 18% and then expect 18% every single day and it's probably going to level out to lower than that. But even if it's 10%, I mean, why would you not want that? Speaker 1: I think it'll make a big difference on my site because we have so many products that are kind of undiscoverable. Because people aren't going to just sort through all the categories, which brings me to the next thing that I fixed over the holidays, or I call it your vacation holiday. Speaker 2: The holiday month of July. Speaker 1: Is I completely fixed my on-site search to use AI as well. Speaker 2: Oh, I was wondering about that. Speaker 1: I also actually just gave a lecture to this to the class, but This took a lot of time. This was like a two, you were gone for how long? Speaker 2: I was gone for a month. Speaker 1: Four weeks. Okay. This one took like two weeks of that four weeks. What I did was I had AI generate very detailed descriptions of all of my products. So it took the photo and described every aspect of that product based on the photo and I had it spit out every possible occasion that that product could be used for, every type of person that'd be looking for it, a whole bunch of things, right? And then I fed that into, I turned that into math with a vector database. And then now when someone queries, that queries that description and returns the closest possible thing. Now, I had neglected my search for a while, so I just kind of looked for the stats. I've been collecting stats about my on-site search for a long time, but I never really, I stopped looking at it after I had it done the first time, which was years ago. And it was something like almost 60% of the people who did searches on mobile linens yielded zero search results. Speaker 2: Interesting. That seems very high. Speaker 1: And the reason for that is because people cannot spell. People use a whole bunch of weird synonyms that aren't spelled correctly. Like Hank, there's actually like 10 ways to say Hanky. Speaker 2: I think handkerchief's a hard word to spell, honestly. Speaker 1: Okay, here, you know what? Here, just give me a sec. Let me bring up... Speaker 2: Also, like you dropped the term Battenberg lace a few minutes ago, and I was like, how do you spell Battenberg? Is it with a E-R-G, a U-R-G? Like, I do think, especially for the types of products that you have, I wonder about that for the curriculum site. Curriculum's a word that gets misspelled all the time. I wonder how often that happens where people are spelling curriculum with one R. I think especially if you have something that can be misspelled a lot of different ways, this is something that's very important. Speaker 1: I brought up all the search results for Bumblebee. Unknown Speaker: How do you spell Hankey? Speaker 2: I or IE? Y? Speaker 1: You can do it all of those, but that's not what I mean. So some people will type long sentences like, I'm looking for a wedding handkerchief. Speaker 2: Yeah. Speaker 1: Out of the box Shopify search is never gonna return anything. Speaker 2: Right. Speaker 1: For that, right? Or people just type in like hydrangeas. Speaker 2: Oh yeah, okay. Speaker 1: And so we don't necessarily have anything for that, but we can show something similar now because AI, it'll return something similar. Speaker 2: Yeah. Speaker 1: Anyway, you get the point, right? Almost 60% of my searches are returning nothing. There's a stat online, and I unfortunately didn't have this tracked, but if someone doesn't get anything in the search results, there's something like almost 70% likely to just leave your site altogether. Speaker 2: Oh, for sure. I believe that. Yeah. Speaker 1: Yeah. Amazon has trained everyone, especially on mobile, to go straight for search. Speaker 2: Yeah. Speaker 1: Which means that I was probably hemorrhaging people. Speaker 2: Mm-hmm. Unknown Speaker: Yeah. Speaker 1: Anyway, so the results are preliminary once again. So this search has been out for three weeks, two weeks? No, how long you been gone? Probably two weeks. And it's already lifted my sales 10% per search. Speaker 2: Yeah. Speaker 1: Those are huge changes. Speaker 2: That makes a lot of sense because I do feel like if you go to a site, you can't find what you want. You immediately just go to the next site. Speaker 1: Right. Speaker 2: Yeah. Speaker 1: Does anyone ever shop on Amazon by looking at categories? I don't think so. Speaker 2: No. No. Never. Speaker 1: I mean, there's probably sometimes, but no. I actually didn't think it would. I thought it would have more of an effect, but I think a lot of my traffic is coming in from ads going to a category page. Speaker 2: Yeah. This is going to help with your organic traffic. Yes, or social or anything if you were on TikTok or something like that and people were just coming in. I have seen, so speaking of all that, since we have been on YouTube, we've seen a huge lift in organic traffic. Nice. People just typing in the name of the store, right? Stuff like that, which I mean, I guess I was expecting, but not to the extent that I'm seeing it, right? So I'm seeing a lot of like that brand recognition. So I'm sure once you get out there on social, this will actually benefit you a whole lot more because people are going to just be typing in the name of the store or coming to you organically. And then that's when they're going to be searching, right? Because they haven't seen a specific product in an ad. Speaker 1: Yeah, yeah. The other thing that I was thinking about doing, and this one requires a little bit more planning, but to create an AI live chat widget, I'm not sure how I feel about it yet because I hate it when I find out someone's AI. But the number of people that are just looking for stuff where you can guide them along, I guess search maybe accomplishes this. Speaker 2: Yeah. Speaker 1: But people also want to know like when is this going to ship or where my order is. Speaker 2: Yeah. Speaker 1: How do you feel about AI chatbots? Speaker 2: So I always hate them because I feel like by the time I get to that, I have a very specific thing that I need, right? Like I don't need to know when it's shipping. I need to know like there's a problem. So that's where I feel like it gets frustrating. But I do like it when I go to a site and I can just get into the bot and get the basic information. You know what I mean? Like, when is it shipping? How do I look at this? Where's the size chart? You know, something like that. Speaker 1: Do you have used it before? Speaker 2: I have and I don't mind it unless it's like, hey, I just want a representative. And then it continues to, when it loops you, when you can't get a representative, that's when I get very frustrated. So that's a tough one, right? Speaker 1: If someone wants a representative, you need to kind of respond pretty quickly, right? So we have someone dedicated for customer support, but I'm just thinking, let's say for one day, just even 10 people want to chat. One person's not going to be able to handle, maybe they can, I don't know. 10 chats though, going out at the same time? Speaker 2: It would almost be like you've got to find a company that does it really well and mimic what they're doing, right? Like, how do you transition people? And I think, you know, one thing that I don't mind is when they tell me that they're a bot, right? When they're like, hey, like to me, if you want to implement something like this on your own site, you have to be very forthcoming with, hey, you're going to get a bot. These are the questions we can answer for you. And then if you need more help, we will move you over to our live customer service kind of thing. I think if that information is at the forefront, then I don't really mind it at all. In fact, I was dealing with that. I think it was the train app that I was using in Europe. Like it was like, hey, we're a bot during these hours, right? So if you need anything, and I was like, okay, that makes like, and then I wasn't on there thinking that I was going to get something that I couldn't get at the time. So I think as long as you're upfront about it, you're probably okay. I don't know. Speaker 1: I just wanted to take a moment to tell you about a free resource that I offer on my website that you may not be aware of. If you are interested in starting your own online store, I put together a comprehensive 6-day mini course on how to get started in ecommerce that you should all check out. It contains both video and text-based tutorials that go over the entire process of finding products to sell all the way to getting your first sales online. Now, this course is free and can be obtained at MyWifeQuitHerJob.com slash free. Just sign up right there on the front page via email and I'll send you the course right away. Once again, that's MyWifeQuitHerJob.com slash free. Now back to the show. Yeah, I mean, it's a hard question. I was kind of checking out how Amazon does it. I mean, they have Rufus and whatnot, but I never really use it. Speaker 2: Yeah, I don't either. Speaker 1: Myself. And yeah, this would, okay, here's what really fascinated me. So Eleven Labs just announced voice customer service. Speaker 2: Oh, interesting. Speaker 1: Yeah. And Google's had this for a long time, where it sounds like a human, it acts like a human. Most people can't tell it's a human, but it's actually voice. Speaker 2: Okay. Speaker 1: And so I have it on my list. I actually already have an 11 Lab subscription and I wanted to try this, but how cool would it be? Well, in my mind at least, how cool would it be to have like a human be able to answer these questions? Speaker 2: Yeah. That's where it gets weird. Speaker 1: That's a harder problem. Yeah. Yeah. Speaker 2: But yeah, I mean, I think I don't know. I think anytime you can minimize the interaction that you have to have, right? If you can get people through a flow that answers their questions, the better, right? Because I'm sure a lot of the questions are stuff that can be answered pretty easily. I mean, there's always going to be the complicated cases, but I don't think that's the norm. It can't be, right? It can't be because why would Amazon do it if it wasn't successful in cutting down? Speaker 1: Well, you'll notice it's not prominent on Amazon. Speaker 2: The Rufus? Speaker 1: Yeah. Speaker 2: Yeah, I guess that's true. Speaker 1: You don't have to look for it, but it's not right front and center for you. So clearly they're testing stuff. Speaker 2: Yeah. Speaker 1: So we used to have a chatbot. I don't know if you remember this. Yeah, I do actually, yeah. I had a chatbot on ManyChat where it answered all those common questions like where's my order, order status, and how long it's going to take to ship. The problem with that was like if you make someone wait, like I would say longer than like even 30 seconds, like they're probably going to just close it and maybe even leave because they think that no one's supporting you, right? That's what I want to avoid. Speaker 2: I like it when you get on the chatbot and you get like a little, I call it like a bubble menu. There's probably an official name, but it's like six bubbles of like, where's my order? I have a problem with my order. My order's damaged. Like, you know, it gives you like some, some basic things and then it's like none of the above and then you click through that. So any of the top ones, it just takes you into those flows. I don't know. Those to me, I don't, I'm not bothered by. But you're right. If you click and then you hear nothing, that I get really irritated about. I'm like, is anybody here? Speaker 1: Right. Actually, this happened with a tool that I paid for just the other day. One, it took them two minutes to even get a response. And then after that, they just gave me a canned, it was a cut and paste. I think it was a human, but they just did a cut and paste. And it actually made me want to unsubscribe from the tool after that. Speaker 2: Yeah. I mean, yeah, I can see that. Speaker 1: Because it was broken. And what I don't want is, it's better to just not even have that than have that experience in my opinion. That's why I actually took off the live chat from Bumblebee. It's no longer there. Speaker 2: Yeah, that's kind of a tough one because it really is about, yeah, that one's tough, I think. Speaker 1: Responsiveness and whatnot. I guess if it got out of hand, I could just hire someone else to handle it. Speaker 2: Do you have a lot of customer service? Speaker 1: We do because it's weddings. Speaker 2: Yeah, and you've got a different demographic. Speaker 1: We have someone answering calls. People like to call us. Speaker 2: Which is crazy to me. Not for your demographics. I can't pay my kids to make a phone call. Speaker 1: But you're on the phone all the time. Speaker 2: I am. I was actually just going to say I was on yesterday with American Express and I'm going to give them a 10 out of 10. Whereas Brian was on a call with Ring Doorbell and they're going to get a zero out of 10. But yeah, here's the thing that I think is important to think before you start implementing any of this stuff, is like my experience with American Express, and it was a Filipino call center, or you know, I could tell that. But my experience was so great, it just increased my loyalty to American Express. And Brian's was so horrid with Ring that I think if we didn't already invest into the Ring system of all the doorbells and all that stuff, he would buy another brand. So I think when you're thinking about customer service type decisions and using AI and bots and stuff like that, I think actually, I feel like we had a talk about this at Sellers Summit one year or we need to have a talk about this is that like customer service can actually be a great sales channel for your business. Based on how people interact with the whole process, right? Like, it can either build that loyalty and like, hey, this is amazing. Like, even if there was a problem initially, it can turn people around versus just like, I mean, literally, I think Brian was ready to rip all the ring doorbells out of my house, right? Like, just because it was so frustrating to get He was in a loop of canned responses. He was on the phone for two hours. Nothing should take two hours. That's not hold. That's talking to people. I think that's something that is really important for your business. Speaker 1: Actually, you know what happened to me recently? I'm going to call this company out because it pissed me off. Terminix, we had a problem a year ago. They didn't even tell me, but they put me on a subscription. I got charged last week hundreds of dollars. They didn't tell me they were putting me on a subscription. I had to call to cancel. You could not cancel online. Speaker 2: I hate that. I hate when you have to call to cancel. Speaker 1: You could not cancel online. It took me 45 minutes to cancel the dang thing. 45 minutes of my day gone. Speaker 2: Oh, that's so frustrating. Anyway, I just think that that's really I think people tend to like try to find the cheapest, easiest customer service solution. And I actually think that's something that can really send your business to another level if you have great customer service. Speaker 1: Oh, no, no, it totally matters. Like when someone calls us, yeah, it's like 80% sale. Speaker 2: Yeah. Speaker 1: Maybe even higher than that. I don't know anymore since I stopped tracking. This is why the decision to do the live chat requires thought and even Jen was a little iffy on that. Anyway, okay, so the next cool thing while you were gone. This is the first thing I implemented actually while you were gone. It was Toni Bot and Steve Bot. Speaker 2: This is so cool. This is the coolest thing you've ever done, honestly. I spent far too much time yesterday with that link that you sent me, reading through. So Steve created this and then he sent me a link yesterday where I could see what people were inputting and it was like a Netflix binge. Speaker 1: Well, you should see Steve Bot. Speaker 2: I wasn't going to ask. I was curious. I was like, I don't want to know. Speaker 1: What's funny is, I don't know, there's some frustrated queries in there sometimes. Speaker 2: Yes, I saw a couple in Tony Bot. Speaker 1: And I don't think people know that I'm tracking, but I did say it before, like I'm tracking all these. Speaker 2: I mean, you have to think if you're typing something in, it's going to get tracked. Like to me, that's sort of a, I don't know, a known thing. But this is, so talk about this, because this is, this is genius, honestly. Speaker 1: So I read in all of the lessons in both of the classes, all the blog posts, all the podcasts, although I haven't put up the podcast just yet, and then all the Sellers Summit videos. Now, you can query them and then you'll get an answer back. It'll find out which lesson that covered that topic and it'll send you a link to that lesson. I was getting a ton of questions that are covering the lessons, but there's like 450 videos in the class now. It's impossible to watch all that. Way better than search. I've noticed people were using my search function on WordPress to find videos. That search function is horrific. Like it doesn't return anything. And so this bot not only does it answer the question without hallucinating because it only grabs from stuff that you've actually said, but it returns an answer and then it tells you what lessons or blog posts or whatever cover that topic. And I think so far people have loved it. Speaker 2: I've gotten a decent amount of feedback. I actually received text. I got a voicemail about it. So I know Toni Bot is very popular. I'm sure Steve Bot is equally as popular. Speaker 1: What's funny was for your bot, I was going to have it flatter me after every response. But then I was thinking to myself, well, what if someone new signs up and they have no idea? So I took that out. But it was fun for a while when I was testing. Speaker 2: I'm sure. I'm sure you had far too much fun with that. I think this is such a great idea because one thing when I was looking through some of the queries yesterday, I think sometimes people don't know exactly like what exactly they want to ask. And so when they put something that's close, it finds the right videos, which I thought was good. And then I also think for you and I, because we made all the videos, And, but I don't remember what the title of that video was, right? So I would have to, so when people, especially when people asked a question that wasn't a, I can answer you in a couple sentences, it was a, I really need to like dig in. I would spend 15 minutes looking through the videos trying to remember where we covered it. Like downloading the PowerPoint slides, trying to find it. So I think it's a huge time saver on both sides. Speaker 1: Yeah, no, totally. I actually even find myself querying it. Like if a student emails me a question and I'm like, I know I covered that, I just can't figure, I can't remember where, I'll just type in to SteveBot and then I'll send them. I need to train people to use SteveBot more. Because it is, I mean, it's new, right? It's like a month old right now at this point. Speaker 2: No, I thought, I think that's a great, that's not, but that's the problem with the Steve and Toni bot is that is something that is, I would say, tougher for people to implement on their own. Like you built, that was, that was a process for you to build that. That wasn't something that someone could just plug and play. Speaker 1: You know what's funny about all this is someone went up to me and they said, hey, maybe you should offer like a service or a plugin or something that does this. It'd be very easy to do a WordPress plugin for this. What's harder is, you know, transcribing the videos and all that stuff. So it could be done. I guess the only downside now is whenever there's a lesson that's added, it needs to go through this whole process now of getting transcribed, generated into the bots and then fed in. I don't know, so many ideas. I'm like debating whether there's like so many plugins and services now with AI that I can easily implement. And I was talking to, who was I talking to this about? Oh, Bernie. I was talking to Bernie about this. I was asking about customer support. Because once I create someone for someone and they pay, it obviously has to work all the time. Whereas right now with the stuff that I'm doing for myself, let's say there's a minor bug. It's not a big deal, so I'll just suck it up and whatever. It's an inconvenience. But once you productize something, you can't do that anymore. So that's always been my dilemma. Speaker 2: What did Bernie have to say? Because he's released products. Speaker 1: He's released, he said customer service wasn't that big of a factor. But then I countered him. I was like, well, to sell your product, I mean, you had to probably answer a ton of questions. He said people are, it's much harder to sell someone a SaaS product, is what he said. Speaker 2: Yeah. Speaker 1: Because one, people just want you to do it for them. Speaker 2: Yeah, I can see that. Speaker 1: And two, people are a little recurring revenue averse now. I've always been. I don't want to pay Adobe a certain amount for something that, well actually I use Adobe every day. That's a bad example. Microsoft. Microsoft Word and all that. I barely use it now to be honest with you. Speaker 2: I use Google Docs for everything. Speaker 1: Exactly. Speaker 2: Or Google Drive or whatever. Speaker 1: Actually, are we done with this? Speaker 2: Yes, yes. Speaker 1: The bot is cool. The final thing that I did is I moved my community over to Discord. Speaker 2: Oh, yes. This is the big one. Speaker 1: That was a huge win. There's people actually just the other day that were like, hey, I don't like how it doesn't have threads. I was like, actually, it does have threads. The problem with the interface is it's kind of complicated in the beginning. As you're getting used to it, but what I love about Discord is that you can code anything into it. So I incorporated SteveBot into Discord so people can just ask, you know, SteveBot questions. Actually, that has not been as popular as I anticipated because then you get to see what other people are querying. Like, everyone can see it. Speaker 2: Yeah, I can see why that's not quite as... Speaker 1: And I think people are a little more squeamish. But, you know, it's still available, whatever. But it's like slack on steroids is the best way for me to describe it. Speaker 2: Yeah, I think I have been in there a little bit. I think it's a better solution for sure. Speaker 1: Oh, it's much better than Facebook groups. Oh, no question. Yeah. Speaker 2: Is there a way to prevent people from being able to see what is queried? Because to me that would be, I don't want people to see what I'm searching. Speaker 1: Yeah, there's probably a way to do it. I could probably create a private SteveBot group that automatically generates and then as soon as you leave it, it dissipates. That's how the voice chat rooms work right now. You double click on it, it creates a room and then you can chat with anyone and as soon as you're both off, the room disappears. I can do that with SteveBot. I question whether it's worth my effort to do that. Because you can just go to the website and use Steve Bot on there privately, right? Speaker 2: Right. So the question about the Discord, because I know you did spend a good amount of time building this out and getting it all set up. Right now it's open to course members. You said you were going to open it up to other people. Speaker 1: Well, not for a little bit. Speaker 2: Are you still thinking that or what's the status? Speaker 1: Well, I am, but it requires infrastructure. Speaker 2: Yeah. Speaker 1: For example, I'm going to have a bot that monitors every message to see if it's promotional. I'm going to have AI give me an analysis. If it's promotional, I'll put that guy in a little penalty box. I haven't written that yet. But these are things that you can't really, well, you certainly can't do it in Facebook. Speaker 2: No, no. Speaker 1: I question whether you can do it in Circle or School or some of these other places to combat spam. I don't know. Speaker 2: Yeah. Speaker 1: But having a spam bot is awesome. Speaker 2: Yeah. Speaker 1: Like if I create the right AI prompts for it and whatnot. Speaker 2: Yeah. Because to me, you're going to want to leverage that, that Discord community. You know what I mean? To help grow the course eventually. Speaker 1: I mean, that's the goal. Yeah. I have too many things like the podcast is launching next week, so many things going on. AI has really got me more excited about the stuff that I've been doing more than a long time, because we've been doing this for a long time. It gets a little old after a while, but all these new possibilities for me have made me really excited lately. Speaker 2: Yeah, it's funny. I was thinking about this over the weekend because I was working on scripting and doing a bunch of things for my new project. And I remember when AI first, when people first started talking about it, you know, two years ago, right? And people were, I think it was even Spencer who did like a hundred blog posts with AI and to see if he could rank. And there was just a lot of stuff like that where I felt like it was kind of garbage, not what Spencer was doing, but in general, like a lot of garbage content being spit out. But now I feel like things have shifted and people are actually using the technology to build, like Liz is building this really cool email tool, right? I think that like, to me, that's the benefit and the interesting part of things. It's definitely made it much more, I don't know, palatable to me, versus initially, I mean, I used it even on my trip. I used it to plot out walking paths and to get safety information, and then I cross-checked it with actual people in the hotel and things like that, because I was like, I'm not going to just trust AI with my own safety, but I don't know, I had it, restaurant reservations, I mean, it was like, I used it a ton just personally, which I think is really cool, and saved me a ton of time, right? Just doing those types of things. Speaker 1: I actually just met up with one of my buddies who's really high up in Google in one of the AI departments. He was telling me, you haven't seen anything yet. I was like, can I work at Google just so I can be exposed? I'm considering going back to work now. I'm actually not even joking about this. Maybe once Karina goes to school, then maybe I'll There's just so much stuff that's happening right now. Remember when the internet first came out? I know we're old enough to remember that. It was really exciting, but it didn't excite me that much to work in industry because you'd be designing routers and hubs, which was, I think, the big thing for me in my industry. But all this AI stuff, we're on the outside right now. We only get what's been tested and released. There's so much stuff under development that We're not privy to and you can only be privy to that stuff if you work at a company that's on the cutting edge of this. Speaker 2: Most people can't do that realistically. Speaker 1: No, that's not true. I think coders and engineers are, I don't know. I think we might be a dying breed once AI allows everyone to do this stuff. Speaker 2: Oh, I'm saying most people can't just go work for a company that's doing those things. Speaker 1: Well, no, they can. What do you mean? Speaker 2: I just couldn't go get a job at Google tomorrow if I wanted to be on the cutting edge. You probably could. Well, I probably could, but I'm just saying most people couldn't. Speaker 1: Yeah, I mean, hiring is really bad right now, obviously. Speaker 2: I don't think anybody can get a job at Google, right? Speaker 1: Yeah, I mean, actually all the tech companies are shedding people because of automation, I think. Speaker 2: Right, so then that leads to, and this is probably a whole other podcast because I know we're out of time. You know, then what happens? I mean, a lot of these software tools you're building and while you're building them today, I feel like in a year from now I can probably build them, right? Yes. So, whereas today I need to pay $15 a month to have access to XYZ, next year I don't because I can just build it myself. Right like yeah, so so then what happens to all the tools all the plugins all the SAS products Right like what? Speaker 1: I don't know what happens so I was We're going to create a video on YouTube about how I think the Shopify app store is in trouble. Speaker 2: We've talked about this for a while. Speaker 1: We talked about this for a while, but I was going to actually pick a bunch of apps that are really popular on the Shopify app store. Speaker 2: You just want to be blackballed by every tool. Speaker 1: Yeah, I thought about that actually because they would be pissed but there's a lot of tools out there. I said like, you know, they don't do anything and they charge you like 20 bucks, 50 bucks a month, right? And it adds up if you don't know how to implement these things. Speaker 2: Oh, I've seen. It's funny because and this is partly from Sellers Summit but like, you know, people will give us access to their accounts to look at things, right? And then you see that They might be on the $79 a month Shopify plan, but their bill every month is $532 because of all of the additional tools that they're paying for. Especially if you're starting a new business, if you're building your brand and you're just starting out, that's a lot of money every month if you're not bringing in that money every month. That's a huge chunk of your expenses. Speaker 1: I mean, people have plugin bloat, to be fair. Like, when you're first starting out, you're not gonna be installing all those plugins. It's only once you start researching, you go, oh, that would be nice. Oh, that would be nice. Speaker 2: I think people get, like, as soon as people do it, I think, haven't you met people that are like, what plugins do I need? I mean, we have people on the course that ask this question a lot, and it's like, well, you don't really need any of them right now. You're not making any sales. But yeah, anyway. Speaker 1: Yeah, that's probably another topic. Speaker 2: Stay tuned on a secret podcast on our Patreon where Steve will list all the apps he thinks are gonna be put out of business in the next 18 months. Speaker 1: It's funny, I get emails from companies who want me to promote them probably every single day. There's this one that recently contacted me and they were like, hey, it's a tool that helps you edit your images and write copy for your product listings. Why the heck would I promote, like why would someone need to pay you to do that stuff, right? When you can just do everything and what's available right now. Speaker 2: Yeah, and that's the one thing that I'm, and this is probably an issue for people that own, that have stores with like actual contractors and employees that work for them, is getting their employees and contractors on board using AI. Because one thing that I have seen is that not everybody that I am dealing with on a regular basis uses it. And I'm like, you could cut your time in half, right? If you used AI to do this part of your job or you used AI to even like summarize some things for you. And that's one thing that I'm seeing and it probably is very industry dependent too. But I think like getting the people that work with you on board and being more efficient is really important because then they have time to do other things that probably will impact your business in a greater way. Speaker 1: Yeah. No, absolutely. So, Toni, when's your next vacation? Speaker 2: I know, right? Next week, actually. Have fun. Next week. Speaker 1: Oh, yeah, that's right. That's right. Okay. Stay tuned for more features. Speaker 2: Stay tuned for what Steve builds when I'm out. Speaker 1: Hope you enjoyed this episode. If you want to know more details about what I've done, which can get a bit technical, feel free to send me email. For more information and resources, go to mywifequitherjob.com slash episode 602. Once again, tickets to the Sellers Summit 2026 are now on sale over at sellerssummit.com. If you want to hang out in person in a small, intimate setting, develop real relationships with like-minded entrepreneurs and learn a ton, then come to my event. Go to SellersSummit.com. And if you're interested in starting your own ecommerce store, head on over to MyWifeQuitHerJob.com and sign up for my free six-day mini course. Just type in your email and I'll send the course right away via email.

This transcript page is part of the Billion Dollar Sellers Content Hub. Explore more content →

Stay Updated

Subscribe to our newsletter to receive updates on new insights and Amazon selling strategies.